Memory – how would you train it? I would do the following:
- Establish metrics
Target:
- Short
- Mid
- Long-term
Modalities:
- Sight
- Concept
- Word
- Sound
- Touch
- Smell
- Etc.
Structure:
- Entity blocks – photographic memory
- Temporal Stream – time-based recall
- Auto-Memory Stream – poetry and recitation until automatic recall
- Associative memory – recall by association
- One to many -> category to contents
- One-to-one/limited -> pairing: language translation
The structure can be also in two modes:
- Detached
- Integrated into existing solid frameworks
So, ideally, I think we should attempt to improve on all permutations of above, but the most valuable is one-limited conceptual recall.
I think we should be aware whether it is possible to rehash memory given that there are physical limitations (limited short-term buffer). It might make sense to explicitly develop the shifting of memory between short<->mid<->long-term buffers.
So I say we set a target and make an exercise to match.
Nikita's Spot 